First, a disclaimer. This commentary is woefully short of sex-related content. Not even the wholesome gender-based outcomes stuff. So, if that key term brought your internet search here, feel free to press on with your illicit journeys my friend. Now, let’s get to the lies.
In 2008, Christopher White, MD, then chair of the Department of Cardiology at the Ochsner Clinic Foundation, published an editorial accusing vascular surgeons of various crimes and misdemeanors in the nascent debate over carotid stenting (CAS) vs. endarterectomy (CEA). Eloquently titled “Liar, liar, pants on fire,” the article begins, “At the most recent Society of Vascular Surgery [SVS] meeting, held in a secret location, deep in a mountain bunker, shielded against critical and independent thinking, surgeons regaled themselves with tales of science fiction, disguised as scientific data, and told to make the endovascularly challenged among them feel safe and secure from the ‘bogeyman’ of carotid artery stenting.” Now, before everyone can post their Spidermen-pointing-at-each-other memes, I will acknowledge the inherent hypocrisy in calling out another medical editor for utilizing hyperbole and provocative titles. So let’s get to the substance of his argument.
The two abstracts that drew White’s ire utilized national databases and showed that CEA demonstrated dramatic benefits over CAS in terms of stroke, mortality and cost. White points out the limitations of the studies, including the lack of independent neurological evaluations to document stroke. He states, “In reporting the superiority of CEA compared to CAS without any qualification, should we give the authors the benefit of the doubt that they are not sophisticated enough to comprehend the inherent ‘bias’ that invalidates their intended comparison, or have we caught them red-handed in an attempt to purposely deceive their audience?” While bemoaning the lack of a limitations section makes me suspect that White does not know how abstracts work, the request for an exam by an independent neurologist seems fair.
So, surely when he published a study in the same issue of this journal, making the provocative claim that carotid artery stenting is safe in the very elderly (>80 years), he held himself to the same lofty standard? After all, the lead-in phase of CREST had just shown a 12% risk of stroke with CAS in octogenarians. White and colleagues, however, reported only a 1.8% risk of stroke in this same age range, even though half were performed without embolic protection. It seems that independent neurologists don’t grow on trees, though, and White’s group often employed vascular medicine physicians from his institution to perform these evaluations, despite no clear evidence that they are either independent or adequately trained to assess for stroke. So perhaps White’s group did achieve these outlier results, or perhaps, as he stated in his editorial, the underreporting of complications is a dirty little secret that is “very difficult to prove without an independent audit.”
White also disputes the finding of higher hospital costs associated with CAS. He states that “the biased data reported in this abstract fly in the face of thousands of reported CAS patients with audited independent data who have had short hospital stays and lower or comparable procedure costs.” I wonder then what he thought of the excellent 2011 analysis by W. Charles Sternbergh, MD, which found that CAS was 40% more costly than CEA and did not provide better clinical outcomes or a reduction in length of stay. I am certain White is aware of this study, as the data came from his own institution.
Over the past 15 years, a plethora of new data has arrived with a series of randomized controlled trials (RCTs). While many used perplexing endpoints (the inclusion of myocardial infarction [MI] seems the logical equivalent of considering ambiance while comparing delivery pizzas), and most were underpowered to detect a statistically significant difference in the risk of stroke, a clear trend emerged. CREST-1 (4.1% vs. 2.3%), ACT-1 (2.8% vs. 1.4%), ICSS (7.7% vs. 4.1%), SAPPHIRE (3.6% vs. 3.1%), SPACE-1 (6.5% vs. 5.1%), EVA-3S (9.2% vs. 3.5%), and ACST-2 (3.6% vs. 2.1%) all demonstrated a persistently increased incidence of 30-day stroke with CAS over CEA. Only CAVATAS was an outlier (7.2% vs. 8.3%), although only 504 patients were randomized.
I was certain that White and his fellow stent enthusiasts would be happy that these new data have brought some much-needed clarity. Surely we can all agree that a good degree of caution must be employed when performing transfemoral carotid stent procedures, particularly in asymptomatic patients where the treatment efficacy is razor-thin? Well reader, I regret to inform you that I was wrong. Last year, the Centers for Medicare & Medicaid Services (CMS) accepted a formal request from the Multispecialty Carotid Alliance to consider greatly expanding coverage for carotid stent procedures while also removing several safeguards currently in place. Although “The Alliance” may sound like the bad guys from a mid-tier Liam Neeson movie, in reality, they are a collection of neurologists, surgeons and interventionalists, including White, who seemingly formed solely to petition CMS to reconsider percutaneous carotid stent coverage.
Medicare began coverage for CAS procedures in 2001 for those performed within investigational device exempt (IDE) studies. In 2005, this expanded to high-risk patients who are symptomatic with greater than 70% lesions, which is essentially where it stands now. Currently, outside of clinical trials, no asymptomatic patients are covered. The last CMS coverage reconsideration for CAS occurred in 2009. According to the Alliance, special interest groups intervened and successfully restricted access to CAS. These “special interest groups” were the SVS, the American Association of Neurological Surgeons, and the American Academy of Neurology.
The Alliance now requests that CMS expand CAS indications to all asymptomatic lesions >70% and all symptomatic lesions >50%. They also request the removal of all facility standards, calling them obsolete. Privileging and credentialing standards, in their plan, will be up to local hospitals. Similarly, all quality assurance would be left to the local peer-review level, eliminating the requirement for participation in a quality registry. Libertarian principles like these are usually better in theory than practice. Allowing folks to police themselves often leads to things like emotional support leopards on domestic flights. Where the evidence does not support expanding CAS indications, the Alliance is quick to find fault. They refer to the stroke/death outcomes in SPACE, EVA-3S, and ICSS as “unacceptable by 2021 standards.” They criticize the inexperience of the CAS operators and the technology and techniques employed at the time. The Alliance would have us believe that EVA-3S was performed in the 1500s by a bunch of French peasants launching stents at unsuspecting patients with trebuchets.
In critiquing the International Carotid Stenting Study (ICSS), presumably because of its 7.7% CAS stroke rate, the Alliance notes that two operators caused five of the strokes despite enrolling only 11 patients. Even though the operators were suspended, the Alliance points out that the strokes were included in the final analysis. To this, I would reply that these practitioners were only removed because the study was carefully monitoring outcomes. How many more real-world strokes are we going to see if guardrails such as registries are removed, as the Alliance recommends? In their submission to CMS, the Alliance states that there are four modern RCTs comparing CAS to CEA in asymptomatic patients. They opine that all have shown comparable outcomes for periprocedural complications, as well as rates of ipsilateral stroke. In reality, none of the protocols were powered to detect a statistically significant difference in stroke outcomes.
The Asymptomatic Carotid Trial (ACT-1) randomized 1,453 patients out of an intended 1,658. The study reported a stroke rate of 2.8% in the stent group and 1.4% in CEA group. Stent-Protected Angioplasty vs. Carotid Endarterectomy-2 (SPACE-2) planned to enroll 3,640 patients but only randomized 513. There were three groups: CEA with best medical therapy (BMT), CAS with BMT, and BMT alone. The five-year cumulative incidence of any ischemic or hemorrhagic stroke was 5.3% in the CEA group, 9.8% with CAS, and 6.5% with BMT. The CAS group also demonstrated the highest (4.4%) five-year cumulative ipsilateral ischemic stroke rate (2% CEA, 3.1% BMT). The Asymptomatic Carotid Surgery Trial-2 (ACST- 2) randomized 3,625 patients. The 30-day stroke rates were higher with CAS (3.6% vs. 2.4%, p=0.06) but did not achieve significance. At five years, the composite risk of ipsilateral stroke or death was 5.5% for CAS and 3.6% for CEA. The Stenting vs. Endarterectomy for Treatment of Carotid-Artery Stenosis (CREST) trial randomized 2,502 patients and did find a significantly higher risk of periprocedural stroke with CAS (4.1% vs. 2.3%, p=0.01). Of course, CREST famously used a combined stroke, MI, and death primary endpoint, which did not demonstrate a significant difference between the modalities. So even though all four modern asymptomatic carotid trials showed a higher incidence of stroke with CAS, the Alliance assures us not to worry. Why? Well, we are better now. Smarter. We have a greatly improved understanding of the anatomic and clinical characteristics that increase risk for CAS. In fact, they even provide them.
So fans of precision medicine take note: Criterion #1— vasculopathy. Vascu-what now? In fact, of the high-risk characteristics provided by the Alliance, the majority are vague definitions like atherosclerotic arch, complex great vessel anatomy and carotid tortuosity. The Alliance correctly touts the importance of patient selection and operator experience in asymptomatic patients, yet they would like to remove all of the guardrails.
So clearly, to offer safe treatment to patients with asymptomatic carotid disease, the practitioner must be technically experienced as well as a content expert. Where will this expertise be gained? Interest does not translate to competence. Vascular surgeons matriculate into the workforce with documented technical and educational experience in the treatment of carotid disease. For other specialties? Not so much. The words “cerebrovascular” and “carotid” appear a total of zero times in the ACGME Program Requirements for Graduate Medical Education in Interventional Cardiology. If CMS follows the recommendations of the Alliance, only one thing is certain. More transfemoral stents will be placed in asymptomatic carotid lesions in Medicare patients by interventionalists under less regulation. The Alliance states that this is in the best interests of patient care. Who is lying now?
Malachi Sheahan III, MD, is the chief medical editor of Vascular Specialist. His opinions do not reflect SVS policy or positions.